[db-wg] New NWI for geofeed?

Tue May 4 22:35:56 CEST 2021

Hello Denis, Colleagues,

Following is the impact analysis for the implementation of the "geofeed:" attribute in the RIPE database, based on the problem statement below and the draft RFC:
https://tools.ietf.org/html/draft-ymbk-opsawg-finding-geofeeds

I will ask our Legal team to conduct a full impact analysis of the implementation plan.

Please reply with corrections or suggestions.

Regards
Ed Shryane
RIPE NCC

Impact Analysis for Implementing the "geofeed:" Attribute 
============================================================

"geoloc:" Attribute
----------------------
Implementing the "geofeed:" attribute does not affect the "geoloc:" attribute. No decision has been taken on the future of the "geoloc:" attribute, a review can be done at a later date.

"remarks:" Attribute
-----------------------
Existing "remarks:" attributes in INETNUM or INET6NUM object types containing a "geofeed: url" value will not be automatically converted to a "geofeed:" attribute.

The implementation will validate that an INETNUM or INET6NUM object may contain at most a single geofeed reference, either a "remarks:" attribute *or* a "geofeed:" attribute. More than one will result in an error on update.

Any "remarks:" attributes in other object types will not be validated for geofeed references.

"geofeed:" Attribute
-----------------------
The "geofeed:" attribute will be added to the INETNUM and INET6NUM object types. It will be an optional, singly occurring attribute.

The attribute value must consist only of a well-formed URL. Any other content in the attribute value will result in a syntax error.

"geofeed:" URL
-----------------
The URL in the "geofeed:" attribute will be validated that it is well-formed (i.e. syntactically correct).

The URL must use the ASCII character set only (in the RIPE database, non-Latin-1 characters will be substituted with a '?' character).

Non-ASCII characters in the URL host name must be converted to ASCII using Punycode in advance (before updating the RIPE database).

Non-ASCII characters in the URL path must be converted using Percent-encoding in advance.

Only the HTTPS protocol is supported in the URL, otherwise an error is returned.

The reachability of the URL will not be checked. The content of the URL will not be validated.

Database dump and Split files
----------------------------------
The "geofeed:" attribute will be included in the nightly database dump and split files.

NRTM
--------
The "geofeed:" attribute will be included in INETNUM and INET6NUM objects in the NRTM stream.

Whois Queries
-----------------
The "geofeed:" attribute will appear by default in (filtered) INETNUM and INET6NUM objects in Whois query responses, no additional query flag will be needed.

RDAP
-------------
The "geofeed:" attribute will not appear in RDAP responses. A separate RDAP profile will be needed to extend the response format to include geofeed. This can be implemented at a later date.

Documentation
---------------
The RIPE database documentation will be updated, including the inet(6)num object templates and attribute description (with a reference to the IETF draft document).

Other RIRs
-------------
There is currently no coordinated plan to implement "geofeed:" across regions. Other RIRs may implement "geofeed:" at a later date.

Legal Review
---------------
An initial review by the RIPE NCC Legal team found that geofeed data may qualify as personal data, and before introducing the "geofeed:" attribute a full impact analysis of its implementation would have to be conducted by the RIPE NCC.

-----

> On 12 Apr 2021, at 17:59, denis walker via db-wg <db-wg at ripe.net> wrote:
> 
> Colleagues
> 
> ** corrected version getting the attribute names right **
> 
> The chairs agree that there is a consensus to set up an NWI to create
> the "geofeed:" attribute in the RIPE Database. We therefore ask the
> RIPE NCC to set up "NWI-13 Create a "geofeed:" attribute in the RIPE
> Database" Using the 'Problem statement' below. After the RIPE NCC
> completes it's impact analysis we can finalise the 'Solution
> definition'. The RIPE NCC can address any of the questions raised in
> this discussion that they feel are relevant to the basic creation of
> this attribute.
> 
> cheers
> denis
> co-chair DB-WG
> 
> 
> Problem statement
> 
> Associating an approximate physical location with an IP address has
> proven to be a challenge to solve within the current constraints of
> the RIPE Database. Over the years the community has chosen to consider
> addresses in the RIPE Database to relate to entities in the assignment
> process itself, not the subsequent actual use of IP addresses after
> assignment.
> 
> The working group is asked to consider whether the RIPE Database can
> be used as a springboard for parties wishing to correlate geographical
> information with IP addresses by allowing structured references in the
> RIPE Database towards information outside the RIPE Database which
> potentially helps answer Geo IP Location queries
> 
> The IETF is currently discussing an update to RPSL to add a new
> attribute "geofeed: url". The url will reference a csv file containing
> location data. Some users have already started to make use of this
> feature via the "remarks: geofeed: url". It is never a good idea to
> try to overload structured data into the free format "remarks:"
> attribute. This has been done in the past, for example with abuse
> contact details before we introduced the "abuse-c:" attribute. There
> is no way to regulate what database users put into "remarks:"
> attributes. So even if the new "geofeed:" attribute is not agreed, the
> url data will still be included in the RIPE Database.
> 
> Currently there are 24,408 INETNUM and 516,354 INET6NUM objects
> containing a "geoloc" attribute in the database. These have 7,731
> distinct values in the INETNUMs and 1,045 distinct values in the
> INET6NUMs. There are about 150 objects in the RIPE Database with a
> "remarks: geoloc url" attribute.
> 
> On Mon, 12 Apr 2021 at 17:56, denis walker <ripedenis at gmail.com> wrote:
>> 
>> Colleagues
>> 
>> The chairs agree that there is a consensus to set up an NWI to create
>> the "geoloc:" attribute in the RIPE Database. We therefore ask the
>> RIPE NCC to set up "NWI-13 Create a "geoloc:" attribute in the RIPE
>> Database" Using the 'Problem statement' below. After the RIPE NCC
>> completes it's impact analysis we can finalise the 'Solution
>> definition'. The RIPE NCC can address any of the questions raised in
>> this discussion that they feel are relevant to the basic creation of
>> this attribute.
>> 
>> cheers
>> denis
>> co-chair DB-WG
>> 
>> 
>> Problem statement
>> 
>> Associating an approximate physical location with an IP address has
>> proven to be a challenge to solve within the current constraints of
>> the RIPE Database. Over the years the community has chosen to consider
>> addresses in the RIPE Database to relate to entities in the assignment
>> process itself, not the subsequent actual use of IP addresses after
>> assignment.
>> 
>> The working group is asked to consider whether the RIPE Database can
>> be used as a springboard for parties wishing to correlate geographical
>> information with IP addresses by allowing structured references in the
>> RIPE Database towards information outside the RIPE Database which
>> potentially helps answer Geo IP Location queries
>> 
>> The IETF is currently discussing an update to RPSL to add a new
>> attribute "geofeed: url". The url will reference a csv file containing
>> location data. Some users have already started to make use of this
>> feature via the "remarks: geofeed: url". It is never a good idea to
>> try to overload structured data into the free format "remarks:"
>> attribute. This has been done in the past, for example with abuse
>> contact details before we introduced the "abuse-c:" attribute. There
>> is no way to regulate what database users put into "remarks:"
>> attributes. So even if the new "geofeed:" attribute is not agreed, the
>> url data will still be included in the RIPE Database.
>> 
>> Currently there are 24,408 INETNUM and 516,354 INET6NUM objects
>> containing a "geoloc:" attribute in the database. These have 7,731
>> distinct values in the INETNUMs and 1,045 distinct values in the
>> INET6NUMs. There are about 150 objects in the RIPE Database with a
>> "remarks: geoloc url" attribute.
>> 
>> On Wed, 7 Apr 2021 at 04:29, denis walker <ripedenis at gmail.com> wrote:
>>> 
>>> HI Massimo
>>> 
>>> I just checked the numbers Ed gave me and I misread the message. These
>>> are the numbers of objects with a "geoloc:" attribute not geofeed :(
>>> 
>>> cheers
>>> denis
>>> co-chair DB-WG
>>> 
>>> On Wed, 7 Apr 2021 at 02:56, Massimo Candela <massimo at us.ntt.net> wrote:
>>>> 
>>>> Hi Denis,
>>>> 
>>>> 
>>>> On 07/04/2021 02:02, denis walker wrote:
>>>>> Your data does not match the data I got from the RIPE NCC...
>>>>> 
>>>>> From the RIPE NCC:
>>>>> 
>>>>> Currently there are 24,408 INETNUM and 516,354 INET6NUM objects
>>>>> containing a "remarks: geofeed: url" attribute in the database. These
>>>>> have 7,731 distinct values in the INETNUMs and 1,045 distinct values
>>>>> in the INET6NUMs.
>>>> 
>>>> 
>>>> I cannot reproduce what you did.
>>>> Even if I just "grep -i geofeed" in ripe.db.inetnum.gz from the ripe ncc
>>>> ftp [1], I obtain only 132 items. And 39 in ripe.db.inet6num.gz. The
>>>> same if I use the complete dump [2].
>>>> 
>>>> Is the data in the FTP wrong? Am I doing something wrong?
>>>> 
>>>> Ciao,
>>>> Massimo
>>>> 
>>>> [1] https://ftp.ripe.net/ripe/dbase/split/
>>>> [2] https://ftp.ripe.net/ripe/dbase/ripe.db.gz
> 